A Unified Framework for Service Availability and Workflow Scheduling in Edge Computing Environment

Abstract

Fog Computing is a new emerging promising paradigm, extending cloud technological resources near to fog edge devices. Fog computing works to give services in highly distributed fashion that has captured recent research and industry trend. Fog platform provides advantage over traditional cloud platform like providing location-based processing, additional intelligence (smart gateways), lower bandwidth consumption, much faster and cheaper to use, real time data delivery and low latency services. The goal of fog computing is not to prove itself better then cloud platform, it just provide real time task and workflow computation better then cloud computing at the network edge that is closer to devices, with that also to make system perform better. The task scheduling and workflow scheduling are two major issues that gained enormous attention and need to be addressed over the fog platform. The task scheduling deals with allocating best fit resources (computation, storage and network resources), data processing and service availability at edge of network. To handle the task scheduling issues, we present a novel architecture for task selection and scheduling at the edge of the network using container as a service (CoaaS). Solving this issue by using cooperative game theory. For that we developed a multi-objective function in order to reduce the energy consumption and makespan by considering different constraints such as memory, CPU, and the users budget. For task selection and scheduling, we have used lightweight containers instead of the conventional virtual machines to reduce the overhead and response time as well as the overall energy consumption of fog devices, that is, nano data centers (nDCs). On other hand scheduling of dynamic dependable multi-task jobs also known as workflow over fog network needs a reliable execution. In the traditional method over cloud network, workflow scheduling deals with various issues like high probability of communication and computing bottlenecks, large data transmission time and failure issues. In addition the smart end users always desire for location-based processing for real time workflows. Again and again accessing the centralized cloud platform to retrieve localized data leading to an inefficient act. So there is need to investigate strategy for reliable and real time workflow scheduling and solving above issues. For thatVMclustering strategy is investigated over edge network, that reduces the execution and scheduling overhead and providing QoS aware workflow scheduling. We present the detailed architecture of workflow scheduling model over Fog- Cloud layer, that integrates the foglets and edge computing platform for better service scheme. We also design an algorithm based on dividing the workflow into sub tasks and finding the most optimal VM cluster for workflow scheduling.